I am analyzing RNASeq data of 18 paired samples (before and after treatment) in frame of DESeq2 pipeline.
In the set some of the pairs are not complete by technical reasons.
I am using straightforward design : model.matrix(~ Ind +Treatment, Exp.Design) ...
So the question is should I omit samples without pair, or still include them?
It seems that DESeqDataSetFromMatrix() and next DESeq() works fine without warnings and errors when samples without pairs are included